US20190251201A1 - Information searching system and information searching method - Google Patents

Information searching system and information searching method Download PDF

Info

Publication number
US20190251201A1
US20190251201A1 US16/140,683 US201816140683A US2019251201A1 US 20190251201 A1 US20190251201 A1 US 20190251201A1 US 201816140683 A US201816140683 A US 201816140683A US 2019251201 A1 US2019251201 A1 US 2019251201A1
Authority
US
United States
Prior art keywords
conversation
history
information
search
unit
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Abandoned
Application number
US16/140,683
Inventor
Kentaro Chigusa
Masashi Tsuchida
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hitachi Ltd
Original Assignee
Hitachi Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hitachi Ltd filed Critical Hitachi Ltd
Assigned to HITACHI, LTD. reassignment HITACHI, LTD. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: CHIGUSA, KENTARO, TSUCHIDA, MASASHI
Publication of US20190251201A1 publication Critical patent/US20190251201A1/en
Abandoned legal-status Critical Current

Links

Images

Classifications

    • G06F17/30696
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/338Presentation of query results
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/3331Query processing
    • G06F16/3332Query translation
    • G06F16/3334Selection or weighting of terms from queries, including natural language queries
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35Clustering; Classification
    • G06F16/353Clustering; Classification into predefined classes
    • G06F17/30663
    • G06F17/30707
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/93Document management systems
    • G06F17/30011

Definitions

  • the present invention relates to an information searching system and an information searching method.
  • Japanese Patent Laid-Open No. 2010-211563 Japanese Patent Laid-Open No. 2010-211563
  • a questioner inputs a question document to a Q&A site of a Q&A system and an answerer inputs an answer document corresponding to the question.
  • a document analysis is multiply performed targeting not only the question document but also the answer document created by the answerer having relatively elaborate knowledge concerning a question event to create a searching index.
  • a searching algorithm is executed using the searching index to perform search processing on a document group included in a database. Consequently, even if a best answer is not directly obtained from the answer document, a useful document latent in the database can be extracted”.
  • Japanese Patent Laid-Open No. 11-143902 it is known that “In a method including a step of storing, as an appearance frequency file, an appearance frequency of a characteristic character string, which is present in a text in a text database, in the text, a step of extracting the characteristic character string from a text designated by a user, and a step of counting an appearance frequency of the characteristic character string in the text designated by the user, a similarity degree to the text designated by the user is calculated using the appearance frequency file and the appearance frequency in the text designated by the user, and a document is searched using the calculated similarity degree”.
  • Japanese Patent Laid-Open No. 2010-9471 Japanese Patent Laid-Open No. 2010-9471
  • a question/answer searching system including a case database in which cases including pairs of question sentences and answer sentences are stored and narrow-down keyword selecting means for calculating importance degrees of keywords included in a question sentence of a case using similarity degrees of the answer sentences paired with the question sentences and selecting, on the basis of the importance degrees, from the keywords, a narrow-down keyword for narrowing down answers to an inquiry text”.
  • an answer document similar to a question sentence is only extracted. Search is not performed considering conversation histories for question sentences present before the question sentence. In the case of relatively simple content, the search sometimes ends through one question and one answer to the question. However, for example, in the case of complicated content or when segmentation of a situation is difficult, an answerer and a questioner need to exchange questions and answers many times.
  • information included in a question sentence in the past sometimes lacks in the latest question sentence.
  • information indicating a background, a prerequisite, or a progress of a question target matter such as a name of a system in which a problem occurs or a situation in which the problem occurs lacks.
  • a person in charge cannot search for an appropriate answer unless the person in charge finds out a history in the past concerning the matter and creates a search query considering the history.
  • the present invention has been devised in view of the above problems, and an object of the present invention is to provide an information searching system and an information searching method that can improve efficiency of information search and convenience of use.
  • an information searching system that searches information
  • the information searching system including: a conversation-history storage unit configured to store conversation histories for inquiries; a conversation-history acquisition unit configured to acquire, from the conversation-history storage unit, a predetermined conversation history corresponding to a search query based on input information; a search-query amendment unit configured to amend the search query on the basis of the predetermined conversation history; and a searching unit configured to search, on the basis of the amended search query, a conversation history stored in the conversation-history storage unit and output a search result.
  • the search query can be amended using the predetermined conversation history corresponding to the search query. Therefore, the search query can be automatically supplemented using the conversation history. As a result, efficiency of information search and convenience of use are improved.
  • FIG. 1 is an overall configuration diagram of a case inquiry system
  • FIG. 2 is a functional configuration diagram of a search execution unit of the case inquiry system
  • FIG. 3 is a flowchart for explaining processing of the search execution unit
  • FIG. 4 is a flowchart for explaining processing of a search-sentence reception unit
  • FIG. 5 is a flowchart for explaining processing of a document searching unit
  • FIG. 6 is a flowchart for explaining processing of a conversation-history acquisition unit
  • FIG. 7 is a flowchart for explaining processing of a query amendment unit
  • FIG. 8 is a flowchart for explaining processing of a conversation-history-text searching unit
  • FIG. 9 is a flowchart for explaining processing of a search-result display unit
  • FIG. 10 is a flowchart for explaining processing of a result reranking unit
  • FIG. 11 is a flowchart for explaining processing of a usefulness-information registration unit
  • FIG. 12 is an explanatory diagram showing the configuration of a conversation-history storage unit
  • FIG. 13 is an explanatory diagram following FIG. 12 ;
  • FIG. 14 is an explanatory diagram showing the configuration of a conversation-history-text storage unit
  • FIG. 15 is an explanatory diagram following FIG. 14 ;
  • FIG. 16 is an explanatory diagram of a search query before amendment
  • FIG. 17 is an explanatory diagram of the search query after amendment
  • FIG. 18 is an explanatory diagram showing the configuration of a handling-code storage unit
  • FIG. 19 is an explanatory diagram showing the configuration of a usefulness-information storage unit
  • FIG. 20 is an explanatory diagram showing the configuration of a user-attribute storage unit
  • FIG. 21 is an explanatory diagram showing a screen before search
  • FIG. 22 is an explanatory diagram showing a screen on which a search result is displayed
  • FIG. 23 is an explanatory diagram showing a screen on which details of the search result are displayed.
  • FIG. 24 is an explanatory diagram showing another screen on which the details of the search result are displayed.
  • FIG. 25 is an explanatory diagram showing the configuration of a handling-code storage unit according to a second embodiment
  • FIG. 26 is a flowchart for explaining processing of a query amendment unit
  • FIG. 27 is an explanatory diagram showing the configuration of a search query after amendment
  • FIG. 28 is an overall configuration diagram of a case searching system according to a third embodiment.
  • FIG. 29 is an overall configuration diagram of a case searching system according to a fourth embodiment.
  • FIG. 30 is an explanatory diagram showing the configuration of a conversation-history storage unit.
  • FIG. 31 is an explanatory diagram showing the configuration of a handling-code storage unit.
  • a conversation history (a behavior history) at the time when a person in charge (an answerer) performs search is acquired and saved.
  • the saved conversation history is used for the next and subsequent search.
  • information exchange between the person in charge and a customer is stored as a conversation history.
  • a search query is amended using the stored conversation history.
  • the conversation history is searched using the amended search query.
  • a conversation history in the past serving as a reference of a problem solution can be efficiently and accurately searched using the search query amended by a conversation history of a present matter. Therefore, according to the embodiments, the number of times of information exchange between the person in charge and the customer can be reduced. A time required for performing the information exchange once can be reduced. The number of handleable matters per unit time of the person in charge can be increased.
  • representations such as “identification information”, “identifier”, “appellation”, “name”, and “ID (IDentification)” are used. These representations can be replaced with one another.
  • not-shown fields may be included, several data fields may be divided and registered in different DBs, or several data fields may be absent.
  • processing steps may be included, several processing steps may be absent, or execution order of several processing steps may be changed.
  • a “computer program” is used as a subject.
  • the computer program is executed by a processor to perform decided processing while using a memory and a communication port (a communication control device). Therefore, the processor may be used as a subject in the explanation.
  • Processing disclosed using the computer program as the subject may be processing performed by a computer or an information processing apparatus such as a management server. Apart or the entire computer program may be realized by dedicated hardware.
  • the program distribution server includes a processor and a storage resource.
  • the storage resource further stores a distribution program and a distribution target computer program.
  • the processor executes the distribution program, whereby the processor of the program distribution server distributes the distribution target computer program to other computers.
  • a part of components in a certain embodiment may be replaced with components in other embodiments.
  • Components in other embodiments may be added to components in a certain embodiment.
  • Other components may be added to, deleted from, or replaced with a part of components of the embodiments.
  • Components, functions, processing units, processing means, and the like explained below may be realized by hardware, such as by designing a part or all of the components, the functions, the processing units, the processing means, and the like as integrated circuits.
  • the components, the functions, the processing units, the processing means, and the like may be realized by software by the processor interpreting and executing computer programs for realizing the functions.
  • Information such as computer programs, tables, and files for realizing the functions only has to be stored in a storage device such as a memory, a hard disk, or an SSD (Solid State Drive) or a recording medium such as an IC card, an SD card, a DVD, a Blue Ray Disk, or another optical disk. Not all of the information needs to be stored in the same recording medium.
  • a storage device such as a memory, a hard disk, or an SSD (Solid State Drive) or a recording medium such as an IC card, an SD card, a DVD, a Blue Ray Disk, or another optical disk.
  • a storage device such as a memory, a hard disk, or an SSD (Solid State Drive) or a recording medium such as an IC card, an SD card, a DVD, a Blue Ray Disk, or another optical disk.
  • IC card an IC card
  • SD card Secure Digital
  • DVD Digital Digital Digital
  • Blue Ray Disk or another optical disk.
  • Control lines and information lines considered to be necessary for explanation are shown. Not all of control lines and information lines necessary for implementation are shown. Actually, almost all the components may be considered to be coupled to one another.
  • the computer includes an input/output device.
  • the input/output device a display, a keyboard, a pointer device, a tablet terminal, and a smartphone are conceivable.
  • the input/output device may be a device other than these devices.
  • a serial interface or an Ethernet interface may be used as an input/output device instead of the input/output device.
  • a computer for display including a display, a keyboard, or a pointer device may be coupled to the interface.
  • Information for display may be transmitted to the computer for display and information for input may be received from the computer for display to perform display in the computer for display.
  • the input and the display in the input/output device may be substituted by receiving an input. Note that the interface is sometimes described as I/F below.
  • the computer includes a communication I/F.
  • a communication I/F As an example of the communication I/F, a LAN (Local Area Network) connection terminal, a SAN (Storage Area Network) connection terminal, and a connecting apparatus of wireless communication are conceivable.
  • the communication I/F may be a device other than these devices.
  • a wired network and a wireless network are conceivable.
  • the communication network may be either the wired network or the wireless network.
  • FIGS. 1 to 24 A first embodiment is explained with reference to FIGS. 1 to 24 . First, definitions of terms used in this embodiment are explained.
  • “Inquiry” means information such as a question sentence or failure information sent from a questioner to a call center or the like.
  • the “inquiry” includes not only a question sent first but also an additional question and information related to questions sent in a series of handling.
  • “Inquiry text” means document information in which content of an inquiry is written.
  • “Handling” means exchanging information for a person in charge, who uses a case searching system, to find a final answer to a question from a questioner, who is a customer.
  • the “handling” includes, for example, confirmation of a situation with the customer, a request for supplementary information, and transfer of the question to a related department.
  • “Matter” means an inquiry unit from the customer. First information exchange to information exchange resulting in problem solution are treated as a set.
  • Mass number means a number uniquely allocated to each matter.
  • Query means information concerning a search expression generated on the basis of a question sentence input to the case searching system by the person in charge.
  • the “query” is referred to as search query as well.
  • Conversation history means history information in which all contents of information exchange between customers (questioners) and persons in charge and related departments concerning inquiries in the past are recorded.
  • Conversation history text means information in which conversation histories are collected as one text file. The “conversation history text” is created for each matter.
  • Machine means a general term of information such as a log or an error message output by a system or an application when a failure occurs.
  • FIG. 1 shows an overall configuration of a case searching system 1 , which is an example of an “information searching system” according to the first embodiment.
  • the case searching system 1 includes, for example, a case searching server 10 , which is an example of a “computer”, an information output apparatus 17 , and an information input apparatus 18 .
  • the case searching system 1 is connected to a terminal 20 of a questioner, who is a customer, via a communication network CN.
  • the questioner is referred to as customer as well.
  • the case searching system 1 is used from, for example, a call center.
  • the case searching system 1 can be provided in a facility such as the call center (so-called on-premise). Alternatively, the case searching system 1 may be disposed on a communication network and configured to be used from one or a plurality of call centers (so-called cloud).
  • the information output apparatus 17 is, for example, a display apparatus.
  • the information output apparatus 17 is not limited to the display apparatus and may include another information output apparatus such as a printer, a sound synthesizing apparatus, or a robot.
  • the information input apparatus 18 is a pointing device such as a keyboard or a mouse.
  • the information input apparatus 18 is not limited to the keyboard and may include another information input apparatus such as a sound recognition apparatus, a visual-line detection apparatus, or an operation detection apparatus.
  • the information output apparatus 17 and the information input apparatus 18 may be integrated like a touch panel.
  • An operation terminal (not shown in FIG. 1 ) coupled to the case searching server 10 can be used as the information output apparatus 17 and the information input apparatus 18 .
  • the case searching system 1 can handle inquiries (referred to as questions as well) from a plurality of customers.
  • a plurality of persons in charge can handle the respective customers using the case searching system 1 .
  • a customer can access the case searching system 1 using the terminal 20 and ask a question via an electronic mail or a chat.
  • the case searching system 1 is not limited to information exchange by characters and may use a voice recognition system to thereby automatically convert voice of the customer or the person in charge into characters and store the characters as an electronic file.
  • a plurality of terminals 20 can simultaneously access the case searching server 10 via the communication network CN.
  • the configuration of the case searching server 10 is explained.
  • the case searching server 10 includes, for example, a microprocessor (CPU) 11 , a memory 12 , an auxiliary storage apparatus (in FIG. 1 , a disk) 13 , a network interface (I/F) 14 , and a storage medium drive 15 .
  • the circuits 11 to 15 are coupled to one another by a bus 16 .
  • the microprocessor 11 executes computer programs explained below stored in the memory 12 to thereby realize functions of the case searching server 10 .
  • the microprocessor 11 uses information T 11 to T 15 explained below stored in the auxiliary storage apparatus 13 .
  • the memory 12 is configured from, for example, a ROM (Read Only Memory) and a RAM (Random Access Memory).
  • ROM Read Only Memory
  • RAM Random Access Memory
  • computer programs such as a search execution unit P 11 , a usefulness-information management unit P 12 , a conversation-history management unit P 13 , and a conversation-history-text registration unit P 14 and an operating system (OS) P 15 are stored.
  • a work area WA used by the microprocessor 11 is also provided.
  • the search execution unit P 11 is a computer program group for performing search.
  • the search execution unit P 11 includes, for example, an inquiry-text reception unit P 111 that receives an inquiry text, a search-result display unit P 112 that causes the customer terminal 20 and/or the information output apparatus 17 to display a search result, and a document searching unit P 113 that searches a conversation history.
  • the document searching unit P 113 includes a query amendment unit P 114 functioning as a “search-query amendment unit” that amends a search query, a conversation-history-text searching unit P 115 that searches a conversation history text, and a result reranking unit P 116 .
  • the conversation-history-text searching unit P 115 is an example of a “searching unit” like a conversation-history searching unit P 115 B explained below with reference to FIG. 28 .
  • the usefulness-information management unit P 12 is a computer program group that manages usefulness information obtained by evaluating whether a conversation history is useful.
  • the usefulness-information management unit P 12 includes, for example, a usefulness-information registration unit P 121 that registers the usefulness information in the case searching system 1 in association with the conversation history and a usefulness-information acquisition unit P 122 that acquires the usefulness information registered in association with the conversation history.
  • the conversation-history management unit P 13 is a computer program group that manages the conversation history.
  • the conversation-history management unit P 13 includes a conversation-history registration unit P 131 that registers the conversation history in the case searching system 1 and a conversation-history acquisition unit P 132 that acquires the registered conversation history.
  • the auxiliary storage apparatus 13 is configured as a nonvolatile large-capacity storage apparatus such as a hard disk or a flash memory device.
  • the auxiliary storage apparatus 13 stores, for example, a conversation-history storage unit T 11 , a handling-code storage unit T 12 , a conversation-history-text storage unit T 13 , a usefulness-information storage unit T 14 , and a user-attribute storage unit T 15 .
  • the conversation-history storage unit T 11 stores the conversation history received from the conversation-history registration unit P 131 .
  • the handling-code storage unit T 12 stores code information associated with the conversation history.
  • the conversation-history-text storage unit T 13 stores the conversation history text generated by the conversation-history-text registration unit P 14 .
  • the usefulness-information storage unit T 14 stores the usefulness information received from the usefulness-information registration unit P 121 .
  • the user-attribute storage unit T 15 stores attributes of a person in charge who answers referring to the conversation history.
  • the network interface 14 is an interface connected to the communication network CN to perform communication.
  • the case searching server 10 is connected from the network interface 14 to the customer terminal 20 via the communication network CN.
  • the storage medium drive 15 is an apparatus that reads and writes an external storage medium MM such as a hard disk or a flash memory.
  • Computer programs P 11 to P 15 or information T 11 to T 15 fixed in the external storage medium MM can be read from the storage medium drive 15 and stored in the memory 12 or the auxiliary storage apparatus 13 .
  • at least a part of the computer programs P 11 to P 15 or the information T 11 to T 15 stored in the memory 12 or the auxiliary storage apparatus 13 can be written and fixed in the external storage medium MM using the storage medium drive 15 .
  • FIG. 2 shows a functional configuration of the search execution unit P 11 .
  • a person in charge inputs an inquiry text to the case searching server 10 from the information input apparatus 18 .
  • the inquiry-text reception unit P 111 of the search execution unit P 11 When receiving the inquiry text (which can be referred to as first searching sentence as well) serving as “input information” from the person in charge, the inquiry-text reception unit P 111 of the search execution unit P 11 passes the inquiry text to the query amendment unit P 114 of the document searching unit P 113 .
  • the conversation-history acquisition unit P 132 acquires a predetermined conversation history corresponding to the inquiry text from the conversation-history storage unit T 11 and acquires code information associated with the predetermined conversation history from the handling-code storage unit T 12 .
  • the conversation-history acquisition unit P 132 sends the acquired predetermined conversation history and the acquired code information to the query amendment unit P 114 .
  • the predetermined conversation history is, for example, a conversation history having the same matter number as a matter number of the inquiry text.
  • the query amendment unit P 114 amends, on the basis of the predetermined conversation history and the code information received from the conversation-history acquisition unit P 132 , a search query (an initial search query) obtained from the input inquiry text. As explained in detail below, the query amendment unit P 114 amends the initial search query by adding the predetermined conversation history to the initial search query and passes the amended search query to the conversation-history-text searching unit P 115 .
  • the conversation-history-text searching unit P 115 searches a conversation history text stored in the conversation-history-text storage unit T 13 using the search query after amendment and outputs a result of the search.
  • the conversation history text is a document obtained by coupling texts of a related plurality of conversation histories as one file.
  • the result reranking unit P 116 functioning as a “search-result rearrangement unit” evaluates the search result of the conversation-history-text searching unit P 115 on the basis of the usefulness information received by the usefulness-information acquisition unit P 122 from the usefulness-information storage unit T 14 and rearranges the conversation histories in descending order of values of the usefulness information.
  • the values of the usefulness information can be changed according to attributes of a user (the person in charge) who evaluates the conversation histories. For example, when an experienced person in charge evaluates that a conversation history is “useful reference”, a value of usefulness information of the conversation history can be set high. Therefore, the usefulness-information acquisition unit P 122 acquires usefulness information for each predetermined conversation history from the usefulness-information storage unit T 14 and acquires attributes (a position, a grade, a skill level, and the like) of the user who sets the usefulness information (the person in charge who refers to the conversation history) from the user-attribute storage unit T 15 .
  • the usefulness-information acquisition unit P 122 can calculate, for example, among the usefulness information set in the predetermined conversation history, the number of pieces of usefulness information set by a user having a predetermined skill level and a predetermined grade and use a value of the number of pieces of usefulness information as a value of the usefulness information.
  • the usefulness-information acquisition unit P 122 can also respectively multiplies the usefulness information set in the predetermined conversation history by values decided from the user attributes and use a total value of the multiplied values as a value of the usefulness information.
  • the search-result display unit P 112 When receiving the search result after the rearrangement from the result reranking unit P 116 , the search-result display unit P 112 transmits the search result to the customer terminal 20 and causes the customer terminal 20 to display the search result.
  • the search-result display unit P 112 may output the search result to the information output apparatus 17 .
  • FIG. 3 is a flowchart for explaining processing of the search execution unit P 11 .
  • the search execution unit P 11 uses the inquiry text received by the inquiry-text reception unit P 111 (S 11 ), causes the document searching unit P 113 to search the conversation history text (S 12 ), and causes the search-result display unit P 112 to display a result of the search (S 13 ).
  • FIG. 4 is a flowchart for explaining processing of the inquiry-text reception unit P 111 .
  • the inquiry-text reception unit P 111 acquires an inquiry text input from the information input apparatus 18 (S 21 ).
  • the inquiry-text reception unit P 111 acquires a matter number of the acquired inquiry text (S 22 ).
  • the person in charge may manually input the matter number together with the inquiry text.
  • the inquiry-text reception unit P 111 may search a conversation history using the inquiry text as a key and acquire a matter number of a conversation history coinciding with the inquiry text. Further, the inquiry-text reception unit P 111 may display a list of matter numbers of conversation histories coinciding with the inquiry text and cause the person in charge to select a matter number out of the matter numbers.
  • the inquiry-text reception unit P 111 creates a query used for first search (an initial search query) from the input inquiry text and the matter number (S 23 ).
  • the inquiry-text reception unit P 111 transmits the created query to the document searching unit P 113 (S 24 ).
  • FIG. 5 is a flowchart for explaining processing of the document searching unit P 113 .
  • the document searching unit P 113 causes the query amendment unit P 114 to amend the initial search query received from the inquiry-text reception unit P 111 (S 31 ).
  • the document searching unit P 113 causes the conversation-history-text searching unit P 115 to search a conversation history text using the query after amendment (S 32 ).
  • the document searching unit P 113 causes the result reranking unit P 116 to rearrange search results on the basis of the values of the usefulness information (S 33 ).
  • FIG. 6 is a flowchart for explaining processing of the conversation-history acquisition unit P 132 .
  • the conversation-history acquisition unit P 132 determines whether a conversation history (a predetermined conversation history) having the same matter number as the matter number of the initial search query is present in the conversation-history storage unit T 11 (S 41 ). That is, the conversation-history acquisition unit P 132 confirms whether a history of a matter having the same matter number as the matter number of the initial search query is stored in the conversation-history storage unit T 11 .
  • a conversation history a predetermined conversation history
  • the conversation-history acquisition unit P 132 acquires a text (a portion excluding bibliographic matters) of the conversation history from the conversation-history storage unit T 11 (S 42 ). If the conversation history having the same matter number as the matter number of the initial search query is not successfully found (NO in S 41 ), this processing ends.
  • FIG. 7 is a flowchart for explaining processing of the query amendment unit P 114 .
  • the query amendment unit P 114 passes the matter number of the initial search query to the conversation-history acquisition unit P 132 and invokes the conversation-history acquisition unit P 132 (S 51 ).
  • the query amendment unit P 114 determines whether the conversation-history acquisition unit P 132 successfully acquires the conversation history having the same matter number (in FIG. 7 , corresponding matter number) as the matter number of the initial search query (S 52 ). If the conversation history having the corresponding matter number is successfully acquired by the conversation-history acquisition unit P 132 (YES in S 52 ), the query amendment unit P 114 acquires a text of the conversation history having the corresponding matter number (S 53 ).
  • the query amendment unit P 114 adds the text acquired in step S 53 to the initial search query to thereby amend the initial search query (S 54 ).
  • the query amendment unit P 114 determines whether another text having the corresponding matter number is still present (S 55 ). If the other text is present (YES in S 55 ), the query amendment unit P 114 returns to step S 53 and further adds the other text to the search query (S 54 ).
  • FIG. 8 is a flowchart for explaining processing of the conversation-history-text searching unit P 115 .
  • the conversation-history-text searching unit P 115 conceptually searches the conversation-history-text storage unit T 13 using the query amended by the query amendment unit P 114 , sorts matter numbers of hit conversation history texts in descending order of scores, and returns the matter numbers (S 61 ).
  • FIG. 9 is a flowchart for explaining processing of the search-result display unit P 112 .
  • the search-result display unit P 112 displays a searching sentence on a screen (S 71 ) and acquires a reranked search result (S 72 ).
  • the search-result display unit P 112 determines whether the next matter number is present in the search result acquired in step S 72 (S 73 ). If the next matter number is absent (NO in S 73 ), the search-result display unit P 112 ends this processing.
  • the search-result display unit P 112 acquires one matter number out of the search result acquired in step S 72 (S 74 ).
  • the search-result display unit P 112 refers to the conversation-history storage unit T 11 and acquires a text of a code “Q1” out of matters coinciding with the matter number acquired in step S 74 (S 75 ).
  • the code “Q1” means “an inquiry from the customer” as shown in FIG. 12 .
  • search-result display unit P 112 refers to the conversation-history storage unit T 11 and acquires a text of a code “AZ” and date and time of the text out of the matters coinciding with the matter number acquired in step S 74 (S 76 ).
  • the code “AZ” means “a final answer to the customer”.
  • the search-result display unit P 112 displays, on the screen, the matter number, the date and time corresponding to the code “AZ”, the text of the code “Q1”, and the text of the code “AZ” (S 77 ).
  • a creation procedure of the conversation-history storage unit T 11 is explained.
  • a person in charge (an answerer) of the call center or the like logs on to the case searching system 1 using a user ID and registers inquiry content in the case searching system 1 .
  • the person in charge sets the content of the inquiry from the customer as a text and registers a handling code (see FIG. 12 ) and a transmission source together.
  • the case searching system 1 invokes the conversation-history registration unit P 131 and cause the conversation-history storage unit T 11 to store a matter number, a progress No., and date and time together.
  • the “matter number” a number obtained by adding “1” to the immediately preceding number is adopted. That is, the conversation-history registration unit P 131 adds “1” to an immediately preceding registered matter number for numbering. 1 is set in the “progress No.”.
  • Registration date and time is set in the “date and time”. The registration date and time can be acquired from a timer in a computer. Note that when the inquiry from the customer is sent by an electronic mail or the like, the inquiry content may be automatically registered in the conversation-history storage unit T 11 as explained above at timing when the electronic mail is received.
  • a creation procedure of the conversation-history-text storage unit T 13 is explained.
  • a text of the conversation-history-text storage unit T 13 is a document obtained by coupling texts having the same matter number among matters stored in the conversation-history storage unit T 11 .
  • a document obtained by coupling texts of a series of handing histories (a conversation history text) is set as a target of a concept search by the search execution unit P 11 .
  • the conversation-history-text registration unit P 14 confirms the conversation-history storage unit T 11 , for example, once in a day.
  • the conversation-history-text registration unit P 14 couples a text of the pertinent matter number and stores the pertinent matter number and the coupled text in the conversation-history-text storage unit T 13 .
  • the conversation-history-text registration unit P 14 may execute the processing at any timing rather than once a day.
  • the conversation-history-text registration unit P 14 may execute the processing, for example, at timing when the code AZ is stored in the conversation-history storage unit T 11 . If a matter currently in progress is also added to a search target, the conversation-history-text registration unit P 14 may be created neglecting a value of the code.
  • the conversation-history-text registration unit P 14 may not be created in advance. That is, information equivalent to the conversation-history-text storage unit T 13 may be created in the work area WA at timing when the search execution unit P 11 executes the concept search.
  • FIG. 10 is a flowchart for explaining processing of the result reranking unit P 116 .
  • the result reranking unit P 116 determines whether one or more search results are present in a result set (S 81 ). If the one or more search results are present (YES in S 81 ), the result reranking unit P 116 proceeds to step S 82 . If the one or more search results are absent (NO in S 81 ), the result reranking unit P 116 proceeds to step S 86 and ends this processing.
  • the result reranking unit P 116 extracts one search result from the result set (S 82 ) and determines whether a matter number of the acquired search result coincides with a matter number stored in the usefulness-information storage unit T 14 (S 83 ). If the matter numbers coincide (YES in S 83 ), the result reranking unit P 116 proceeds to step S 84 . Otherwise (NO in S 83 ), the result reranking unit P 116 returns to step S 81 .
  • the result reranking unit P 116 acquires a registrant ID corresponding to the matter number of the usefulness-information storage unit T 14 , searches the user-attribute storage unit T 15 concerning the acquired registrant ID, and determines whether a role of the registrant is a leader (S 84 ). If the registrant is the “leader” (YES in S 84 ), the result reranking unit P 116 proceeds to step S 85 . Otherwise, the result reranking unit P 116 returns to step S 82 .
  • the result reranking unit P 116 may determine the registrant not only according to the role condition but also according to a plurality of conditions connected by AND or OR. Instead of the “role”, a numerical value such as a “level” may be registered in the user-attribute storage unit T 15 . The result reranking unit P 116 may determine, for example, whether “the level is equal to or larger than three”.
  • the result reranking unit P 116 increments a score of the search result by one (S 85 ) and returns to step S 82 .
  • the result reranking unit P 116 sorts a search result set in descending order of corrected scores (S 86 ).
  • FIG. 11 is a flowchart for explaining processing of the usefulness-information registration unit P 121 .
  • the usefulness-information registration unit P 121 starts at timing when a “useful reference” button B 31 of a case searching system screen example (detailed display) G 1 ( 3 ) shown in FIG. 23 is pressed.
  • the usefulness-information registration unit P 121 acquires a matter number displayed on the search result screen (detailed display) G 1 ( 3 ) (S 91 ) and acquires a user ID of a user who presses the button B 31 (S 91 ).
  • the usefulness-information registration unit P 121 causes the usefulness-information storage unit T 14 to store the acquired matter number and the acquired user ID (S 93 ).
  • FIGS. 12 and 13 show an example of storage content of the conversation-history storage unit T 11 .
  • the conversation-history storage unit T 11 manages, for example, “matter number”, “progress No.”, “date and time”, “transmission source”, “code”, and “text” in association with one another.
  • the “transmission source” is a person in charge (in FIG. 12 , “call center”) or a customer.
  • FIGS. 14 and 15 show an example of storage content of the conversation-history-text storage unit T 13 .
  • a conversation history text is generated by coupling texts of conversation histories having the same matter number.
  • FIG. 16 shows an example of a search query before amendment.
  • FIG. 17 shows an example of a search query after amendment.
  • the search query before amendment includes, as shown on the lower side of FIG. 17 , only information provision (a code QL) from a customer.
  • the search query after amendment further includes, as shown on the upper side of FIG. 17 , an inquiry (a code Q1) from the customer and an information provision request (a code AL) from a person in charge to the customer.
  • the search query is amended by adding information extracted from a series of information in the past of the case.
  • FIG. 18 shows an example of storage content of the handling-code storage unit T 12 . Codes are prepared in advance to classify content of a conversation history.
  • a code Q1 indicates an inquiry from a customer.
  • a code AL indicates an information provision request to the customer.
  • a code QL indicates information provision from the customer.
  • a code AX indicates contact with the customer.
  • a code QX indicates supplementary information from the customer.
  • a code Q2 indicates an additional question from the customer.
  • a code EQ indicates escalation to a related department. The escalation means an inquiry to the related department.
  • a code EA indicates an answer from the related department.
  • a code AZ indicates a final answer to the customer.
  • FIG. 19 shows an example of storage content of the usefulness-information storage unit T 14 .
  • the usefulness-information storage unit T 14 stores, for example, a matter number and a registrant ID in association with each other.
  • FIG. 20 shows an example of storage content of the user-attribute storage unit T 15 .
  • the user-attribute storage unit T 15 stores, for example, a user ID and a role in association with each other.
  • the roll is information indicating a role of a user who uses information concerning a case in the past (a person in charge who answers a question).
  • the role is classified into, for example, “leader” and “person in charge”.
  • attributes such as “subleader”, “supervisor”, and “observer” may be added.
  • FIG. 21 shows an example before search of a case searching screen G 1 ( 1 ) for searching a case.
  • the screen G 1 ( 1 ) includes, for example, a section GP 11 ( 1 ) for displaying an inquiry text and a search button B 11 .
  • FIG. 22 shows an example after search of a case searching screen G 1 ( 2 ).
  • the screen G 1 ( 2 ) after a search result includes, for example, an inquiry-text display section GP 11 ( 1 ), a search button B 11 , and a search result section GP 12 .
  • an inquiry-text display section GP 11 ( 1 ) includes, for example, an inquiry-text display section GP 11 ( 1 ), a search button B 11 , and a search result section GP 12 .
  • FIG. 23 shows an example of a detailed display screen G 1 ( 3 ) of the case searching screen.
  • a list of search results is displayed on the screen G 1 ( 2 ) shown in FIG. 22
  • the person in charge clicks a link set in the desired matter number
  • a detailed progress of the matter is displayed.
  • a detailed screen may be directly provided to the person in charge. That is, the person in charge does not need to perform operation for, for example, clicking a link. In this case, for example, the person in charge has to display only a highest-order matter of a search result.
  • the screen G 1 ( 3 ) shown in FIG. 23 includes, for example, a final answer section GP 11 ( 2 ) for displaying a final answer, a “useful reference” button B 31 , and a progress display section GP 12 ( 2 ).
  • a final answer from the person in charge (a text of the code AZ) is displayed.
  • an inquiry text from the customer may be displayed or the final answer section GP 11 ( 2 ) itself may be omitted from the screen.
  • all progresses may be displayed as shown in FIG. 23 or only a part of the progresses may be displayed as shown in FIG. 24 .
  • a progress in or after date and time including a text coinciding most with an inquiry text used for search is displayed.
  • the code AL at “2017/9/1 10:50:00” corresponds to the progress. Because a part of the progresses is displayed, the person in charge can easily understand the next action that the person in charge should perform for the customer in response to an inquiry from the customer. In the case of FIGS. 23 and 24 , the person in charge understand that an action to be performed next is to request the customer to send a log file (server.log). By automatically extracting the action to be performed next, the case searching system 1 can also be applied to an automatic response system.
  • the search query can be amended according to a predetermined conversation history corresponding to the search query. Therefore, according to this embodiment, the search query can be automatically supplemented using a conversation history. Efficiency of information search and convenience of use are improved.
  • the code information for classifying the content (handling content) of the conversation history for each of types is used. Therefore, the conversation history can be managed according to the content.
  • the conversation history used for amendment of the search query can be easily extracted.
  • a second embodiment is explained with reference to FIGS. 25 to 27 .
  • Embodiments explained below including this embodiment are equivalent to modifications of the first embodiment. Therefore, differences from the first embodiment are mainly explained.
  • weight is given in advance to each of codes set in a conversation history. The weight is given to a search query to perform search.
  • FIG. 25 is a storage example of a handling-code storage unit T 12 A.
  • the handling-code storage unit T 12 A stores a code, a meaning of the code, and weight of the code in association with one another.
  • Relatively high weight is set for information given from a customer and relatively low weight is set for information given from a person in charge to the customer. This is because the information (a text) given from the person in charge is considered to be not very useful for solution of a problem of the customer.
  • FIG. 26 is a flowchart for explaining processing of a query amendment unit P 114 A.
  • the processing shown in FIG. 26 is different from the processing shown in FIG. 7 in that weight is acquired and added to a search query together with a text.
  • the query amendment unit P 114 A passes a matter number of an initial search query to the conversation-history acquisition unit P 132 and invokes the conversation-history acquisition unit P 132 (S 51 ).
  • the query amendment unit P 114 A determines whether the conversation-history acquisition unit P 132 successfully acquires a conversation history and weight having the same matter number as the matter number of the initial search query (S 52 A).
  • the query amendment unit P 114 A acquire a text and weight of the conversation history having the corresponding matter number (S 53 A).
  • the query amendment unit P 114 A adds the text and the weight acquired in step S 53 A to the initial search query to thereby amend the initial search query (S 54 A).
  • the query amendment unit P 114 A determines whether another text (and weight) having the corresponding matter number is still present (S 55 A). If the other text (and weight) is present (YES in S 55 A), the query amendment unit P 114 A returns to step S 53 A and further adds the other text to the search query (S 54 A).
  • FIG. 27 shows an example of a search query after amendment. This embodiment is different from the first embodiment in that weight is given to each of codes in the search query. The weight can be reflected on search as explained below.
  • the scores are given to the words extracted from the document of the query after amendment in this way.
  • a conversation history text is conceptually searched using the words with the scores. Note that when a word appears a plurality of times like the “HB-2330” described above, only a word having a highest score has to be left.
  • This embodiment configured as explained above has the same operational effects as the operational effects in the first embodiment. Further, according to this embodiment, because the weight can be given to each of the codes, the conversation history text can be conceptually searched preferentially using a keyword (a word) included in important handling. A more appropriate search result can be obtained and convenience of use is improved.
  • a third embodiment is explained with reference to FIG. 28 .
  • a case searching system 1 B of this embodiment is different from the case searching system 1 of the first embodiment in that the case searching system 1 B performs the concept search on texts of conversation histories rather than performing the concept search on the conversation history text.
  • the conversation history text is conceptually searched using the search query after amendment (the amended search query).
  • a search target in this embodiment is texts of a conversation history.
  • a conversation-history searching unit P 115 B of a case searching server 10 B of this embodiment conceptually searches texts of conversation histories with the search query after amendment.
  • the conversation-history searching unit P 115 B outputs a result of the search in matter number units. Processing of the result reranking unit P 116 and the search-result display unit P 112 is the same as the processing in the first embodiment.
  • This embodiment configured as explained above has the same operational effects as the operational effects in the first embodiment.
  • the texts of the conversation histories are directly conceptually searched without using a conversation history text obtained by coupling the texts of the conversation histories. Therefore, the configuration of the case searching server 10 B can be simplified.
  • This embodiment can be combined with the second embodiment and used.
  • a fourth embodiment is explained with reference to FIGS. 29 to 31 .
  • a case searching system 1 C is applied to a chatbot system (a system for automatically responding to an inquiry).
  • FIG. 29 is a configuration example of the case searching system 1 C.
  • a case searching server 10 C includes an inquiry reception unit P 111 C instead of the inquiry-text reception unit P 111 . Further, the case searching server 10 C includes an answering unit P 112 C instead of the search-result display unit P 112 .
  • a person in charge in a call center does not answer an inquiry from a customer. Instead, the case searching system 1 C functions as a chatbot and automatically answers the inquiry from the customer.
  • the inquiry reception unit P 111 C of this embodiment receives an inquiry from a customer.
  • the answering unit P 112 C of this embodiment outputs, as a chat, a text of a code ZZ included in a highest-order matter number of a result of reranking by the result reranking unit P 116 .
  • FIG. 30 shows an example of a conversation-history storage unit T 11 C.
  • the code ZZ is a text created for an automatic answer on the basis of handling of a corresponding matter number.
  • the text of the code ZZ is registered by a person in charge in a call center or a separately assigned answer creator by invoking the conversation-history registration unit P 131 from the case searching system 1 C.
  • the text of the code ZZ is created by excluding a customer name, a person in charge name, a greeting phrase, and the like from the text of the code AZ. Besides, the text of the code ZZ may be created anew for automatic response or may be cited from a manual or the like corresponding to the text.
  • FIG. 31 shows an example of a handling-code storage unit T 12 C.
  • the code ZZ indicates an answer for automatic response as explained above.
  • the case searching system 1 C can be incorporated in a chat system and can automatically answer an inquiry from a customer. Therefore, the case searching system 1 C can be efficiently operated.

Abstract

An information searching system 1 includes a conversation-history storage unit T11 configured to store conversation histories for inquiries, a conversation-history acquisition unit P132 configured to acquire, from the conversation-history storage unit T11, a predetermined conversation history corresponding to a search query based on input information, a search-query amendment unit P114 configured to amend the search query on the basis of the predetermined conversation history, and a searching unit P115 configured to search, on the basis of the amended search query, the conversation histories stored in the conversation-history storage unit T11 and output a search result.

Description

    BACKGROUND
  • The present invention relates to an information searching system and an information searching method.
  • In a call center or the like that answers inquiries from customers, customer handling results are recorded. When a new inquiry is received, an answerer searches similar cases in the past and answers the inquiry referring to a result of the search. For the search of cases in the past, besides full-text search for searching a case including a designated keyword, concept search for searching a case including a document similar to an inquiry text is used.
  • As one related art (Japanese Patent Laid-Open No. 2010-211563), it is known that “A questioner inputs a question document to a Q&A site of a Q&A system and an answerer inputs an answer document corresponding to the question. Thereafter, a document analysis is multiply performed targeting not only the question document but also the answer document created by the answerer having relatively elaborate knowledge concerning a question event to create a searching index. A searching algorithm is executed using the searching index to perform search processing on a document group included in a database. Consequently, even if a best answer is not directly obtained from the answer document, a useful document latent in the database can be extracted”.
  • As another related art (Japanese Patent Laid-Open No. 11-143902), it is known that “In a method including a step of storing, as an appearance frequency file, an appearance frequency of a characteristic character string, which is present in a text in a text database, in the text, a step of extracting the characteristic character string from a text designated by a user, and a step of counting an appearance frequency of the characteristic character string in the text designated by the user, a similarity degree to the text designated by the user is calculated using the appearance frequency file and the appearance frequency in the text designated by the user, and a document is searched using the calculated similarity degree”.
  • As still another related art (Japanese Patent Laid-Open No. 2010-9471), there is known “A question/answer searching system including a case database in which cases including pairs of question sentences and answer sentences are stored and narrow-down keyword selecting means for calculating importance degrees of keywords included in a question sentence of a case using similarity degrees of the answer sentences paired with the question sentences and selecting, on the basis of the importance degrees, from the keywords, a narrow-down keyword for narrowing down answers to an inquiry text”.
  • SUMMARY
  • In the related arts, an answer document similar to a question sentence is only extracted. Search is not performed considering conversation histories for question sentences present before the question sentence. In the case of relatively simple content, the search sometimes ends through one question and one answer to the question. However, for example, in the case of complicated content or when segmentation of a situation is difficult, an answerer and a questioner need to exchange questions and answers many times.
  • When information such as question sentences and answer sentences is exchanged a plurality of times between the questioner and the answerer, information included in a question sentence in the past sometimes lacks in the latest question sentence. For example, information indicating a background, a prerequisite, or a progress of a question target matter such as a name of a system in which a problem occurs or a situation in which the problem occurs lacks. In this case, a person in charge cannot search for an appropriate answer unless the person in charge finds out a history in the past concerning the matter and creates a search query considering the history.
  • The present invention has been devised in view of the above problems, and an object of the present invention is to provide an information searching system and an information searching method that can improve efficiency of information search and convenience of use.
  • In order to solve the problems, an information searching system according to an aspect of the present invention is an information searching system that searches information, the information searching system including: a conversation-history storage unit configured to store conversation histories for inquiries; a conversation-history acquisition unit configured to acquire, from the conversation-history storage unit, a predetermined conversation history corresponding to a search query based on input information; a search-query amendment unit configured to amend the search query on the basis of the predetermined conversation history; and a searching unit configured to search, on the basis of the amended search query, a conversation history stored in the conversation-history storage unit and output a search result.
  • According to the present invention, the search query can be amended using the predetermined conversation history corresponding to the search query. Therefore, the search query can be automatically supplemented using the conversation history. As a result, efficiency of information search and convenience of use are improved.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • FIG. 1 is an overall configuration diagram of a case inquiry system;
  • FIG. 2 is a functional configuration diagram of a search execution unit of the case inquiry system;
  • FIG. 3 is a flowchart for explaining processing of the search execution unit;
  • FIG. 4 is a flowchart for explaining processing of a search-sentence reception unit;
  • FIG. 5 is a flowchart for explaining processing of a document searching unit;
  • FIG. 6 is a flowchart for explaining processing of a conversation-history acquisition unit;
  • FIG. 7 is a flowchart for explaining processing of a query amendment unit;
  • FIG. 8 is a flowchart for explaining processing of a conversation-history-text searching unit;
  • FIG. 9 is a flowchart for explaining processing of a search-result display unit;
  • FIG. 10 is a flowchart for explaining processing of a result reranking unit;
  • FIG. 11 is a flowchart for explaining processing of a usefulness-information registration unit;
  • FIG. 12 is an explanatory diagram showing the configuration of a conversation-history storage unit;
  • FIG. 13 is an explanatory diagram following FIG. 12;
  • FIG. 14 is an explanatory diagram showing the configuration of a conversation-history-text storage unit;
  • FIG. 15 is an explanatory diagram following FIG. 14;
  • FIG. 16 is an explanatory diagram of a search query before amendment;
  • FIG. 17 is an explanatory diagram of the search query after amendment;
  • FIG. 18 is an explanatory diagram showing the configuration of a handling-code storage unit;
  • FIG. 19 is an explanatory diagram showing the configuration of a usefulness-information storage unit;
  • FIG. 20 is an explanatory diagram showing the configuration of a user-attribute storage unit;
  • FIG. 21 is an explanatory diagram showing a screen before search;
  • FIG. 22 is an explanatory diagram showing a screen on which a search result is displayed;
  • FIG. 23 is an explanatory diagram showing a screen on which details of the search result are displayed;
  • FIG. 24 is an explanatory diagram showing another screen on which the details of the search result are displayed;
  • FIG. 25 is an explanatory diagram showing the configuration of a handling-code storage unit according to a second embodiment;
  • FIG. 26 is a flowchart for explaining processing of a query amendment unit;
  • FIG. 27 is an explanatory diagram showing the configuration of a search query after amendment;
  • FIG. 28 is an overall configuration diagram of a case searching system according to a third embodiment;
  • FIG. 29 is an overall configuration diagram of a case searching system according to a fourth embodiment;
  • FIG. 30 is an explanatory diagram showing the configuration of a conversation-history storage unit; and
  • FIG. 31 is an explanatory diagram showing the configuration of a handling-code storage unit.
  • DETAILED DESCRIPTION OF THE EMBODIMENTS
  • Embodiments of the present invention are explained below with reference to the drawings. In the embodiments, as explained below, a conversation history (a behavior history) at the time when a person in charge (an answerer) performs search is acquired and saved. The saved conversation history is used for the next and subsequent search.
  • That is, in the embodiments, information exchange between the person in charge and a customer (a questioner) is stored as a conversation history. When the conversation history is searched, a search query is amended using the stored conversation history. The conversation history is searched using the amended search query.
  • Consequently, according to the embodiments, a conversation history in the past serving as a reference of a problem solution can be efficiently and accurately searched using the search query amended by a conversation history of a present matter. Therefore, according to the embodiments, the number of times of information exchange between the person in charge and the customer can be reduced. A time required for performing the information exchange once can be reduced. The number of handleable matters per unit time of the person in charge can be increased.
  • In the following explanation, information in the embodiments is explained using representations such as “aaa table”, “aaa list”, and “aaa DB (Database)” (aaa is any character string). However, these kinds of information do not always need to be saved in the forms and may be represented in forms other than data structures such as a table, a list, a DB, and a queue. Therefore, the “aaa table”, the “aaa list”, the “aaa DB”, and the like are sometimes referred to as “aaa information” in order to indicate that the information does not depend on the data structure.
  • When contents of the information are explained, representations such as “identification information”, “identifier”, “appellation”, “name”, and “ID (IDentification)” are used. These representations can be replaced with one another.
  • In the information illustrated in the embodiments, not-shown fields may be included, several data fields may be divided and registered in different DBs, or several data fields may be absent.
  • In processing flows illustrated in the embodiments, not-shown processing steps may be included, several processing steps may be absent, or execution order of several processing steps may be changed.
  • In the following explanation, a “computer program” is used as a subject. However, the computer program is executed by a processor to perform decided processing while using a memory and a communication port (a communication control device). Therefore, the processor may be used as a subject in the explanation. Processing disclosed using the computer program as the subject may be processing performed by a computer or an information processing apparatus such as a management server. Apart or the entire computer program may be realized by dedicated hardware.
  • Various computer programs may be installed in computers by a program distribution server or a computer-readable storage medium. In this case, the program distribution server includes a processor and a storage resource. The storage resource further stores a distribution program and a distribution target computer program. The processor executes the distribution program, whereby the processor of the program distribution server distributes the distribution target computer program to other computers.
  • Note that the present invention is not limited to the embodiments explained below and includes various modifications and equivalent configurations within the gist of the appended claims. For example, the embodiments are explained in detail in order to clearly explain the present invention. The embodiments are not always limited to embodiments including all explained components.
  • A part of components in a certain embodiment may be replaced with components in other embodiments. Components in other embodiments may be added to components in a certain embodiment. Other components may be added to, deleted from, or replaced with a part of components of the embodiments.
  • Components, functions, processing units, processing means, and the like explained below may be realized by hardware, such as by designing a part or all of the components, the functions, the processing units, the processing means, and the like as integrated circuits. The components, the functions, the processing units, the processing means, and the like may be realized by software by the processor interpreting and executing computer programs for realizing the functions.
  • Information such as computer programs, tables, and files for realizing the functions only has to be stored in a storage device such as a memory, a hard disk, or an SSD (Solid State Drive) or a recording medium such as an IC card, an SD card, a DVD, a Blue Ray Disk, or another optical disk. Not all of the information needs to be stored in the same recording medium. Several computer programs, tables, and files may be stored in different storage media or may be stored in storage media of not-shown apparatuses connected via a communication network.
  • Control lines and information lines considered to be necessary for explanation are shown. Not all of control lines and information lines necessary for implementation are shown. Actually, almost all the components may be considered to be coupled to one another.
  • The computer includes an input/output device. As examples of the input/output device, a display, a keyboard, a pointer device, a tablet terminal, and a smartphone are conceivable. However, the input/output device may be a device other than these devices. A serial interface or an Ethernet interface may be used as an input/output device instead of the input/output device. A computer for display including a display, a keyboard, or a pointer device may be coupled to the interface. Information for display may be transmitted to the computer for display and information for input may be received from the computer for display to perform display in the computer for display. The input and the display in the input/output device may be substituted by receiving an input. Note that the interface is sometimes described as I/F below.
  • The computer includes a communication I/F. As an example of the communication I/F, a LAN (Local Area Network) connection terminal, a SAN (Storage Area Network) connection terminal, and a connecting apparatus of wireless communication are conceivable. However, the communication I/F may be a device other than these devices. As the communication network, a wired network and a wireless network are conceivable. The communication network may be either the wired network or the wireless network.
  • First Embodiment
  • A first embodiment is explained with reference to FIGS. 1 to 24. First, definitions of terms used in this embodiment are explained.
  • “Inquiry” means information such as a question sentence or failure information sent from a questioner to a call center or the like. The “inquiry” includes not only a question sent first but also an additional question and information related to questions sent in a series of handling.
  • “Inquiry text” means document information in which content of an inquiry is written.
  • “Handling” means exchanging information for a person in charge, who uses a case searching system, to find a final answer to a question from a questioner, who is a customer. The “handling” includes, for example, confirmation of a situation with the customer, a request for supplementary information, and transfer of the question to a related department.
  • “Matter” means an inquiry unit from the customer. First information exchange to information exchange resulting in problem solution are treated as a set.
  • “Matter number” means a number uniquely allocated to each matter.
  • “Query” means information concerning a search expression generated on the basis of a question sentence input to the case searching system by the person in charge. The “query” is referred to as search query as well.
  • “Case” means a completed matter.
  • “Conversation history” means history information in which all contents of information exchange between customers (questioners) and persons in charge and related departments concerning inquiries in the past are recorded.
  • “Conversation history text” means information in which conversation histories are collected as one text file. The “conversation history text” is created for each matter.
  • “Material” means a general term of information such as a log or an error message output by a system or an application when a failure occurs.
  • FIG. 1 shows an overall configuration of a case searching system 1, which is an example of an “information searching system” according to the first embodiment.
  • The case searching system 1 includes, for example, a case searching server 10, which is an example of a “computer”, an information output apparatus 17, and an information input apparatus 18. The case searching system 1 is connected to a terminal 20 of a questioner, who is a customer, via a communication network CN. In the following explanation, the questioner is referred to as customer as well. The case searching system 1 is used from, for example, a call center. The case searching system 1 can be provided in a facility such as the call center (so-called on-premise). Alternatively, the case searching system 1 may be disposed on a communication network and configured to be used from one or a plurality of call centers (so-called cloud).
  • The information output apparatus 17 is, for example, a display apparatus. The information output apparatus 17 is not limited to the display apparatus and may include another information output apparatus such as a printer, a sound synthesizing apparatus, or a robot. The information input apparatus 18 is a pointing device such as a keyboard or a mouse. The information input apparatus 18 is not limited to the keyboard and may include another information input apparatus such as a sound recognition apparatus, a visual-line detection apparatus, or an operation detection apparatus. The information output apparatus 17 and the information input apparatus 18 may be integrated like a touch panel.
  • An operation terminal (not shown in FIG. 1) coupled to the case searching server 10 can be used as the information output apparatus 17 and the information input apparatus 18. The case searching system 1 can handle inquiries (referred to as questions as well) from a plurality of customers. A plurality of persons in charge can handle the respective customers using the case searching system 1.
  • A customer can access the case searching system 1 using the terminal 20 and ask a question via an electronic mail or a chat. The case searching system 1 is not limited to information exchange by characters and may use a voice recognition system to thereby automatically convert voice of the customer or the person in charge into characters and store the characters as an electronic file. Although not shown in FIG. 1, a plurality of terminals 20 can simultaneously access the case searching server 10 via the communication network CN.
  • The configuration of the case searching server 10 is explained. The case searching server 10 includes, for example, a microprocessor (CPU) 11, a memory 12, an auxiliary storage apparatus (in FIG. 1, a disk) 13, a network interface (I/F) 14, and a storage medium drive 15. The circuits 11 to 15 are coupled to one another by a bus 16.
  • The microprocessor 11 executes computer programs explained below stored in the memory 12 to thereby realize functions of the case searching server 10. In executing the computer program, the microprocessor 11 uses information T11 to T15 explained below stored in the auxiliary storage apparatus 13.
  • The memory 12 is configured from, for example, a ROM (Read Only Memory) and a RAM (Random Access Memory). In the memory 12, for example, computer programs such as a search execution unit P11, a usefulness-information management unit P12, a conversation-history management unit P13, and a conversation-history-text registration unit P14 and an operating system (OS) P15 are stored. In the memory 12, a work area WA used by the microprocessor 11 is also provided.
  • The search execution unit P11 is a computer program group for performing search. The search execution unit P11 includes, for example, an inquiry-text reception unit P111 that receives an inquiry text, a search-result display unit P112 that causes the customer terminal 20 and/or the information output apparatus 17 to display a search result, and a document searching unit P113 that searches a conversation history.
  • The document searching unit P113 includes a query amendment unit P114 functioning as a “search-query amendment unit” that amends a search query, a conversation-history-text searching unit P115 that searches a conversation history text, and a result reranking unit P116. The conversation-history-text searching unit P115 is an example of a “searching unit” like a conversation-history searching unit P115B explained below with reference to FIG. 28.
  • The usefulness-information management unit P12 is a computer program group that manages usefulness information obtained by evaluating whether a conversation history is useful. The usefulness-information management unit P12 includes, for example, a usefulness-information registration unit P121 that registers the usefulness information in the case searching system 1 in association with the conversation history and a usefulness-information acquisition unit P122 that acquires the usefulness information registered in association with the conversation history.
  • The conversation-history management unit P13 is a computer program group that manages the conversation history. The conversation-history management unit P13 includes a conversation-history registration unit P131 that registers the conversation history in the case searching system 1 and a conversation-history acquisition unit P132 that acquires the registered conversation history.
  • The auxiliary storage apparatus 13 is configured as a nonvolatile large-capacity storage apparatus such as a hard disk or a flash memory device. The auxiliary storage apparatus 13 stores, for example, a conversation-history storage unit T11, a handling-code storage unit T12, a conversation-history-text storage unit T13, a usefulness-information storage unit T14, and a user-attribute storage unit T15.
  • The conversation-history storage unit T11 stores the conversation history received from the conversation-history registration unit P131. The handling-code storage unit T12 stores code information associated with the conversation history. The conversation-history-text storage unit T13 stores the conversation history text generated by the conversation-history-text registration unit P14. The usefulness-information storage unit T14 stores the usefulness information received from the usefulness-information registration unit P121. The user-attribute storage unit T15 stores attributes of a person in charge who answers referring to the conversation history.
  • The network interface 14 is an interface connected to the communication network CN to perform communication. The case searching server 10 is connected from the network interface 14 to the customer terminal 20 via the communication network CN.
  • The storage medium drive 15 is an apparatus that reads and writes an external storage medium MM such as a hard disk or a flash memory. Computer programs P11 to P15 or information T11 to T15 fixed in the external storage medium MM can be read from the storage medium drive 15 and stored in the memory 12 or the auxiliary storage apparatus 13. Alternatively, at least a part of the computer programs P11 to P15 or the information T11 to T15 stored in the memory 12 or the auxiliary storage apparatus 13 can be written and fixed in the external storage medium MM using the storage medium drive 15.
  • FIG. 2 shows a functional configuration of the search execution unit P11. When receiving an inquiry from a customer through information transmitting means such as an electronic mail or a chat, a person in charge inputs an inquiry text to the case searching server 10 from the information input apparatus 18.
  • When receiving the inquiry text (which can be referred to as first searching sentence as well) serving as “input information” from the person in charge, the inquiry-text reception unit P111 of the search execution unit P11 passes the inquiry text to the query amendment unit P114 of the document searching unit P113.
  • According to a request from the query amendment unit P114, the conversation-history acquisition unit P132 acquires a predetermined conversation history corresponding to the inquiry text from the conversation-history storage unit T11 and acquires code information associated with the predetermined conversation history from the handling-code storage unit T12. The conversation-history acquisition unit P132 sends the acquired predetermined conversation history and the acquired code information to the query amendment unit P114. The predetermined conversation history is, for example, a conversation history having the same matter number as a matter number of the inquiry text.
  • The query amendment unit P114 amends, on the basis of the predetermined conversation history and the code information received from the conversation-history acquisition unit P132, a search query (an initial search query) obtained from the input inquiry text. As explained in detail below, the query amendment unit P114 amends the initial search query by adding the predetermined conversation history to the initial search query and passes the amended search query to the conversation-history-text searching unit P115.
  • The conversation-history-text searching unit P115 searches a conversation history text stored in the conversation-history-text storage unit T13 using the search query after amendment and outputs a result of the search. The conversation history text is a document obtained by coupling texts of a related plurality of conversation histories as one file.
  • The result reranking unit P116 functioning as a “search-result rearrangement unit” evaluates the search result of the conversation-history-text searching unit P115 on the basis of the usefulness information received by the usefulness-information acquisition unit P122 from the usefulness-information storage unit T14 and rearranges the conversation histories in descending order of values of the usefulness information.
  • The values of the usefulness information can be changed according to attributes of a user (the person in charge) who evaluates the conversation histories. For example, when an experienced person in charge evaluates that a conversation history is “useful reference”, a value of usefulness information of the conversation history can be set high. Therefore, the usefulness-information acquisition unit P122 acquires usefulness information for each predetermined conversation history from the usefulness-information storage unit T14 and acquires attributes (a position, a grade, a skill level, and the like) of the user who sets the usefulness information (the person in charge who refers to the conversation history) from the user-attribute storage unit T15.
  • The usefulness-information acquisition unit P122 can calculate, for example, among the usefulness information set in the predetermined conversation history, the number of pieces of usefulness information set by a user having a predetermined skill level and a predetermined grade and use a value of the number of pieces of usefulness information as a value of the usefulness information. Alternatively, the usefulness-information acquisition unit P122 can also respectively multiplies the usefulness information set in the predetermined conversation history by values decided from the user attributes and use a total value of the multiplied values as a value of the usefulness information.
  • When receiving the search result after the rearrangement from the result reranking unit P116, the search-result display unit P112 transmits the search result to the customer terminal 20 and causes the customer terminal 20 to display the search result. The search-result display unit P112 may output the search result to the information output apparatus 17.
  • FIG. 3 is a flowchart for explaining processing of the search execution unit P11. As explained above, the search execution unit P11 uses the inquiry text received by the inquiry-text reception unit P111 (S11), causes the document searching unit P113 to search the conversation history text (S12), and causes the search-result display unit P112 to display a result of the search (S13).
  • FIG. 4 is a flowchart for explaining processing of the inquiry-text reception unit P111. The inquiry-text reception unit P111 acquires an inquiry text input from the information input apparatus 18 (S21). The inquiry-text reception unit P111 acquires a matter number of the acquired inquiry text (S22).
  • The person in charge may manually input the matter number together with the inquiry text. Alternatively, the inquiry-text reception unit P111 may search a conversation history using the inquiry text as a key and acquire a matter number of a conversation history coinciding with the inquiry text. Further, the inquiry-text reception unit P111 may display a list of matter numbers of conversation histories coinciding with the inquiry text and cause the person in charge to select a matter number out of the matter numbers.
  • The inquiry-text reception unit P111 creates a query used for first search (an initial search query) from the input inquiry text and the matter number (S23). The inquiry-text reception unit P111 transmits the created query to the document searching unit P113 (S24).
  • FIG. 5 is a flowchart for explaining processing of the document searching unit P113. The document searching unit P113 causes the query amendment unit P114 to amend the initial search query received from the inquiry-text reception unit P111 (S31).
  • The document searching unit P113 causes the conversation-history-text searching unit P115 to search a conversation history text using the query after amendment (S32). The document searching unit P113 causes the result reranking unit P116 to rearrange search results on the basis of the values of the usefulness information (S33).
  • FIG. 6 is a flowchart for explaining processing of the conversation-history acquisition unit P132. The conversation-history acquisition unit P132 determines whether a conversation history (a predetermined conversation history) having the same matter number as the matter number of the initial search query is present in the conversation-history storage unit T11 (S41). That is, the conversation-history acquisition unit P132 confirms whether a history of a matter having the same matter number as the matter number of the initial search query is stored in the conversation-history storage unit T11.
  • If finding the conversation history having the same matter number as the matter number of the initial search query (YES in S41), the conversation-history acquisition unit P132 acquires a text (a portion excluding bibliographic matters) of the conversation history from the conversation-history storage unit T11 (S42). If the conversation history having the same matter number as the matter number of the initial search query is not successfully found (NO in S41), this processing ends.
  • FIG. 7 is a flowchart for explaining processing of the query amendment unit P114. The query amendment unit P114 passes the matter number of the initial search query to the conversation-history acquisition unit P132 and invokes the conversation-history acquisition unit P132 (S51).
  • The query amendment unit P114 determines whether the conversation-history acquisition unit P132 successfully acquires the conversation history having the same matter number (in FIG. 7, corresponding matter number) as the matter number of the initial search query (S52). If the conversation history having the corresponding matter number is successfully acquired by the conversation-history acquisition unit P132 (YES in S52), the query amendment unit P114 acquires a text of the conversation history having the corresponding matter number (S53).
  • The query amendment unit P114 adds the text acquired in step S53 to the initial search query to thereby amend the initial search query (S54). The query amendment unit P114 determines whether another text having the corresponding matter number is still present (S55). If the other text is present (YES in S55), the query amendment unit P114 returns to step S53 and further adds the other text to the search query (S54).
  • If the conversation history having the corresponding matter number is unsuccessfully acquired (NO in S52) and if a text (another text) of another conversation history having the corresponding matter number is absent (NO in S55), the processing ends.
  • FIG. 8 is a flowchart for explaining processing of the conversation-history-text searching unit P115. For example, the conversation-history-text searching unit P115 conceptually searches the conversation-history-text storage unit T13 using the query amended by the query amendment unit P114, sorts matter numbers of hit conversation history texts in descending order of scores, and returns the matter numbers (S61).
  • FIG. 9 is a flowchart for explaining processing of the search-result display unit P112. The search-result display unit P112 displays a searching sentence on a screen (S71) and acquires a reranked search result (S72).
  • The search-result display unit P112 determines whether the next matter number is present in the search result acquired in step S72 (S73). If the next matter number is absent (NO in S73), the search-result display unit P112 ends this processing.
  • If determining that the next matter number is present (YES in S73), the search-result display unit P112 acquires one matter number out of the search result acquired in step S72 (S74). The search-result display unit P112 refers to the conversation-history storage unit T11 and acquires a text of a code “Q1” out of matters coinciding with the matter number acquired in step S74 (S75). The code “Q1” means “an inquiry from the customer” as shown in FIG. 12.
  • Further, the search-result display unit P112 refers to the conversation-history storage unit T11 and acquires a text of a code “AZ” and date and time of the text out of the matters coinciding with the matter number acquired in step S74 (S76). The code “AZ” means “a final answer to the customer”.
  • The search-result display unit P112 displays, on the screen, the matter number, the date and time corresponding to the code “AZ”, the text of the code “Q1”, and the text of the code “AZ” (S77).
  • A creation procedure of the conversation-history storage unit T11 is explained.
  • (Procedure 1) A conversation history is created in a job of a call center or the like.
  • (Procedure 2) When receiving a first report of an inquiry from a customer, a person in charge (an answerer) of the call center or the like logs on to the case searching system 1 using a user ID and registers inquiry content in the case searching system 1. For example, the person in charge sets the content of the inquiry from the customer as a text and registers a handling code (see FIG. 12) and a transmission source together.
  • (Procedure 3) The case searching system 1 invokes the conversation-history registration unit P131 and cause the conversation-history storage unit T11 to store a matter number, a progress No., and date and time together. As the “matter number”, a number obtained by adding “1” to the immediately preceding number is adopted. That is, the conversation-history registration unit P131 adds “1” to an immediately preceding registered matter number for numbering. 1 is set in the “progress No.”. Registration date and time is set in the “date and time”. The registration date and time can be acquired from a timer in a computer. Note that when the inquiry from the customer is sent by an electronic mail or the like, the inquiry content may be automatically registered in the conversation-history storage unit T11 as explained above at timing when the electronic mail is received.
  • (Procedure 4) When the person in charge answers the customer, confirms details with the customer, requests the customer to send materials, and sends an inquiry to a related department such as a development source department of an inquiry target product, the person in charge stores an inquiry text and a code in the conversation-history storage unit T11 using the case searching system 1 every time. In that case, the progress No. automatically increases by one at a time.
  • A creation procedure of the conversation-history-text storage unit T13 is explained.
  • (Procedure 1) A text of the conversation-history-text storage unit T13 is a document obtained by coupling texts having the same matter number among matters stored in the conversation-history storage unit T11. A document obtained by coupling texts of a series of handing histories (a conversation history text) is set as a target of a concept search by the search execution unit P11.
  • (Procedure 2) The conversation-history-text registration unit P14 confirms the conversation-history storage unit T11, for example, once in a day. When a matter number not stored in the conversation-history-text storage unit T13 is present and the code “AZ” is present in a matter of the matter number (i.e., a final answer to the matter is obtained), the conversation-history-text registration unit P14 couples a text of the pertinent matter number and stores the pertinent matter number and the coupled text in the conversation-history-text storage unit T13.
  • (Procedure 3) Note that the conversation-history-text registration unit P14 may execute the processing at any timing rather than once a day. The conversation-history-text registration unit P14 may execute the processing, for example, at timing when the code AZ is stored in the conversation-history storage unit T11. If a matter currently in progress is also added to a search target, the conversation-history-text registration unit P14 may be created neglecting a value of the code.
  • (Procedure 4) The conversation-history-text registration unit P14 may not be created in advance. That is, information equivalent to the conversation-history-text storage unit T13 may be created in the work area WA at timing when the search execution unit P11 executes the concept search.
  • (Procedure 5) In the conversation history text, not only exchange of information between the person in charge and the customer but also a technical document such as a manual or a document such as a collection of knowhow may be registered together and set as a search target. Consequently, the person in charge can access more accurate information. As a result, a handling time further decreases.
  • FIG. 10 is a flowchart for explaining processing of the result reranking unit P116. First, the result reranking unit P116 determines whether one or more search results are present in a result set (S81). If the one or more search results are present (YES in S81), the result reranking unit P116 proceeds to step S82. If the one or more search results are absent (NO in S81), the result reranking unit P116 proceeds to step S86 and ends this processing.
  • The result reranking unit P116 extracts one search result from the result set (S82) and determines whether a matter number of the acquired search result coincides with a matter number stored in the usefulness-information storage unit T14 (S83). If the matter numbers coincide (YES in S83), the result reranking unit P116 proceeds to step S84. Otherwise (NO in S83), the result reranking unit P116 returns to step S81.
  • The result reranking unit P116 acquires a registrant ID corresponding to the matter number of the usefulness-information storage unit T14, searches the user-attribute storage unit T15 concerning the acquired registrant ID, and determines whether a role of the registrant is a leader (S84). If the registrant is the “leader” (YES in S84), the result reranking unit P116 proceeds to step S85. Otherwise, the result reranking unit P116 returns to step S82.
  • Note that the result reranking unit P116 may determine the registrant not only according to the role condition but also according to a plurality of conditions connected by AND or OR. Instead of the “role”, a numerical value such as a “level” may be registered in the user-attribute storage unit T15. The result reranking unit P116 may determine, for example, whether “the level is equal to or larger than three”.
  • The result reranking unit P116 increments a score of the search result by one (S85) and returns to step S82.
  • As a result of performing the processing in this way, when it is finally determined “NO” in step S81, the result reranking unit P116 sorts a search result set in descending order of corrected scores (S86).
  • FIG. 11 is a flowchart for explaining processing of the usefulness-information registration unit P121. The usefulness-information registration unit P121 starts at timing when a “useful reference” button B31 of a case searching system screen example (detailed display) G1(3) shown in FIG. 23 is pressed.
  • The usefulness-information registration unit P121 acquires a matter number displayed on the search result screen (detailed display) G1(3) (S91) and acquires a user ID of a user who presses the button B31 (S91). The usefulness-information registration unit P121 causes the usefulness-information storage unit T14 to store the acquired matter number and the acquired user ID (S93).
  • FIGS. 12 and 13 show an example of storage content of the conversation-history storage unit T11. The conversation-history storage unit T11 manages, for example, “matter number”, “progress No.”, “date and time”, “transmission source”, “code”, and “text” in association with one another. In FIG. 12, the “transmission source” is a person in charge (in FIG. 12, “call center”) or a customer.
  • FIGS. 14 and 15 show an example of storage content of the conversation-history-text storage unit T13. A conversation history text is generated by coupling texts of conversation histories having the same matter number.
  • FIG. 16 shows an example of a search query before amendment.
  • FIG. 17 shows an example of a search query after amendment. The search query before amendment includes, as shown on the lower side of FIG. 17, only information provision (a code QL) from a customer. On the other hand, the search query after amendment further includes, as shown on the upper side of FIG. 17, an inquiry (a code Q1) from the customer and an information provision request (a code AL) from a person in charge to the customer. In this way, in this embodiment, rather than searching a case in the past according to only the latest inquiry from the customer, the search query is amended by adding information extracted from a series of information in the past of the case.
  • FIG. 18 shows an example of storage content of the handling-code storage unit T12. Codes are prepared in advance to classify content of a conversation history.
  • A code Q1 indicates an inquiry from a customer. A code AL indicates an information provision request to the customer. A code QL indicates information provision from the customer. A code AX indicates contact with the customer. A code QX indicates supplementary information from the customer. A code Q2 indicates an additional question from the customer. A code EQ indicates escalation to a related department. The escalation means an inquiry to the related department. A code EA indicates an answer from the related department. A code AZ indicates a final answer to the customer. These codes are an example. Codes having other meanings may be added. A part of the codes shown in FIG. 18 may be deleted.
  • FIG. 19 shows an example of storage content of the usefulness-information storage unit T14. The usefulness-information storage unit T14 stores, for example, a matter number and a registrant ID in association with each other.
  • FIG. 20 shows an example of storage content of the user-attribute storage unit T15. The user-attribute storage unit T15 stores, for example, a user ID and a role in association with each other. The roll is information indicating a role of a user who uses information concerning a case in the past (a person in charge who answers a question). The role is classified into, for example, “leader” and “person in charge”. Besides, for example, attributes such as “subleader”, “supervisor”, and “observer” may be added.
  • FIG. 21 shows an example before search of a case searching screen G1(1) for searching a case. The screen G1(1) includes, for example, a section GP11(1) for displaying an inquiry text and a search button B11.
  • FIG. 22 shows an example after search of a case searching screen G1(2). The screen G1(2) after a search result includes, for example, an inquiry-text display section GP11(1), a search button B11, and a search result section GP12. When the person in charge selects a desired matter number, as shown in FIG. 23 or FIG. 24, detailed information is presented.
  • FIG. 23 shows an example of a detailed display screen G1(3) of the case searching screen. In this embodiment, after a list of search results is displayed on the screen G1(2) shown in FIG. 22, when the person in charge (the user) clicks a link set in the desired matter number, a detailed progress of the matter is displayed.
  • Instead of this, a detailed screen may be directly provided to the person in charge. That is, the person in charge does not need to perform operation for, for example, clicking a link. In this case, for example, the person in charge has to display only a highest-order matter of a search result.
  • The screen G1(3) shown in FIG. 23 includes, for example, a final answer section GP11(2) for displaying a final answer, a “useful reference” button B31, and a progress display section GP12(2). In the final answer section GP11(2), a final answer from the person in charge (a text of the code AZ) is displayed. Instead of this, an inquiry text from the customer may be displayed or the final answer section GP11(2) itself may be omitted from the screen.
  • In the progress display section GP12(2), all progresses may be displayed as shown in FIG. 23 or only a part of the progresses may be displayed as shown in FIG. 24.
  • In the screen G1(4) shown in FIG. 24, for example, a progress in or after date and time including a text coinciding most with an inquiry text used for search is displayed. In the examples shown in FIGS. 23 and 24, the code AL at “2017/9/1 10:50:00” corresponds to the progress. Because a part of the progresses is displayed, the person in charge can easily understand the next action that the person in charge should perform for the customer in response to an inquiry from the customer. In the case of FIGS. 23 and 24, the person in charge understand that an action to be performed next is to request the customer to send a log file (server.log). By automatically extracting the action to be performed next, the case searching system 1 can also be applied to an automatic response system.
  • According to this embodiment configured as explained above, the search query can be amended according to a predetermined conversation history corresponding to the search query. Therefore, according to this embodiment, the search query can be automatically supplemented using a conversation history. Efficiency of information search and convenience of use are improved.
  • According to this embodiment, the code information for classifying the content (handling content) of the conversation history for each of types is used. Therefore, the conversation history can be managed according to the content. The conversation history used for amendment of the search query can be easily extracted.
  • Second Embodiment
  • A second embodiment is explained with reference to FIGS. 25 to 27. Embodiments explained below including this embodiment are equivalent to modifications of the first embodiment. Therefore, differences from the first embodiment are mainly explained. In this embodiment, weight is given in advance to each of codes set in a conversation history. The weight is given to a search query to perform search.
  • FIG. 25 is a storage example of a handling-code storage unit T12A. The handling-code storage unit T12A stores a code, a meaning of the code, and weight of the code in association with one another.
  • Relatively high weight is set for information given from a customer and relatively low weight is set for information given from a person in charge to the customer. This is because the information (a text) given from the person in charge is considered to be not very useful for solution of a problem of the customer.
  • FIG. 26 is a flowchart for explaining processing of a query amendment unit P114A. The processing shown in FIG. 26 is different from the processing shown in FIG. 7 in that weight is acquired and added to a search query together with a text.
  • The query amendment unit P114A passes a matter number of an initial search query to the conversation-history acquisition unit P132 and invokes the conversation-history acquisition unit P132 (S51). The query amendment unit P114A determines whether the conversation-history acquisition unit P132 successfully acquires a conversation history and weight having the same matter number as the matter number of the initial search query (S52A).
  • If the conversation history having the corresponding matter number is successfully acquired by the conversation-history acquisition unit P132 (YES in S52A), the query amendment unit P114A acquire a text and weight of the conversation history having the corresponding matter number (S53A).
  • The query amendment unit P114A adds the text and the weight acquired in step S53A to the initial search query to thereby amend the initial search query (S54A). The query amendment unit P114A determines whether another text (and weight) having the corresponding matter number is still present (S55A). If the other text (and weight) is present (YES in S55A), the query amendment unit P114A returns to step S53A and further adds the other text to the search query (S54A).
  • If the conversation history having the corresponding matter number is not successfully acquired (NO in S52A) and if a text (the other text) of another conversation history having the corresponding matter number is absent (NO in S55A), this processing ends.
  • FIG. 27 shows an example of a search query after amendment. This embodiment is different from the first embodiment in that weight is given to each of codes in the search query. The weight can be reflected on search as explained below.
  • It is assumed that words and scores of the words (in parentheses) extracted from a text “Backup of yesterday did not end within the job stop time at night in the server HB-2330 of the inhouse system of DEF securities” to which a code Q1 is given are HB-2330 (100), backup (40), and did not end (10). In this case, by multiplying the extracted scores by weight “1.0” set for the code Q1, the scores of the words in this query are HB-2330 (100), backup (40), and did not end (10).
  • Similarly, it is assumed that words and scores of the words extracted from a text “Is an error message shown? Please check the message file abc.err and let me know. *** Saito, Support Service, bbb company ***” to which a code AL is given are error message (100), message file (30), check (20), and in charge (30). In this case, by multiplying the extracted scores by weight “0.2” set for the code AL, the scores of the words in this query change to HB-2330 (20), message file (6), check (4), and in charge (6). The same applies to a text to which a code QL is given.
  • In this embodiment, the scores are given to the words extracted from the document of the query after amendment in this way. A conversation history text is conceptually searched using the words with the scores. Note that when a word appears a plurality of times like the “HB-2330” described above, only a word having a highest score has to be left.
  • This embodiment configured as explained above has the same operational effects as the operational effects in the first embodiment. Further, according to this embodiment, because the weight can be given to each of the codes, the conversation history text can be conceptually searched preferentially using a keyword (a word) included in important handling. A more appropriate search result can be obtained and convenience of use is improved.
  • Third Embodiment
  • A third embodiment is explained with reference to FIG. 28. A case searching system 1B of this embodiment is different from the case searching system 1 of the first embodiment in that the case searching system 1B performs the concept search on texts of conversation histories rather than performing the concept search on the conversation history text.
  • That is, in the first embodiment, the conversation history text is conceptually searched using the search query after amendment (the amended search query). However, a search target in this embodiment is texts of a conversation history. A conversation-history searching unit P115B of a case searching server 10B of this embodiment conceptually searches texts of conversation histories with the search query after amendment. The conversation-history searching unit P115B outputs a result of the search in matter number units. Processing of the result reranking unit P116 and the search-result display unit P112 is the same as the processing in the first embodiment.
  • This embodiment configured as explained above has the same operational effects as the operational effects in the first embodiment. In this embodiment, the texts of the conversation histories are directly conceptually searched without using a conversation history text obtained by coupling the texts of the conversation histories. Therefore, the configuration of the case searching server 10B can be simplified. This embodiment can be combined with the second embodiment and used.
  • Fourth Embodiment
  • A fourth embodiment is explained with reference to FIGS. 29 to 31. In this embodiment, a case searching system 1C is applied to a chatbot system (a system for automatically responding to an inquiry).
  • FIG. 29 is a configuration example of the case searching system 1C. A case searching server 10C includes an inquiry reception unit P111C instead of the inquiry-text reception unit P111. Further, the case searching server 10C includes an answering unit P112C instead of the search-result display unit P112. In this embodiment, a person in charge in a call center does not answer an inquiry from a customer. Instead, the case searching system 1C functions as a chatbot and automatically answers the inquiry from the customer.
  • The inquiry reception unit P111C of this embodiment receives an inquiry from a customer. The answering unit P112C of this embodiment outputs, as a chat, a text of a code ZZ included in a highest-order matter number of a result of reranking by the result reranking unit P116.
  • FIG. 30 shows an example of a conversation-history storage unit T11C. As indicated by thick characters in the lower side of FIG. 30, the code ZZ is a text created for an automatic answer on the basis of handling of a corresponding matter number. The text of the code ZZ is registered by a person in charge in a call center or a separately assigned answer creator by invoking the conversation-history registration unit P131 from the case searching system 1C.
  • The text of the code ZZ is created by excluding a customer name, a person in charge name, a greeting phrase, and the like from the text of the code AZ. Besides, the text of the code ZZ may be created anew for automatic response or may be cited from a manual or the like corresponding to the text.
  • FIG. 31 shows an example of a handling-code storage unit T12C. The code ZZ indicates an answer for automatic response as explained above.
  • This embodiment configured as explained above has the same operational effects as the operational effects in the first embodiment. Further, according to this embodiment, the case searching system 1C can be incorporated in a chat system and can automatically answer an inquiry from a customer. Therefore, the case searching system 1C can be efficiently operated.

Claims (9)

What is claimed is:
1. An information searching system that searching information, the information searching system comprising:
a conversation-history storage unit configured to store conversation histories for inquiries;
a conversation-history acquisition unit configured to acquire, from the conversation-history storage unit, a predetermined conversation history corresponding to a search query based on input information;
a search-query amendment unit configured to amend the search query on the basis of the predetermined conversation history; and
a searching unit configured to search, on the basis of the amended search query, the conversation histories stored in the conversation-history storage unit and output a search result.
2. The information searching system according to claim 1, wherein
the search query has a matter number given to each of inquiry matters,
the conversation-history storage unit is configured to associate the matter number and handling content in past to thereby store the matter number and the handling content as the conversation histories,
the conversation-history acquisition unit is configured to acquire a conversation history having a same matter number as the matter number of the search query from the conversation-history storage unit as the predetermined conversation history, and
the search-query amendment unit is configured to add at least a part of handling content included in the predetermined conversation history to the search query to thereby amend the search query.
3. The information searching system according to claim 2, wherein the conversation history includes code information for classifying the handling content for each of types.
4. The information searching system according to claim 3, wherein
weight information is set in the code information, and
the search-query amendment unit is configured to add handling content included in the predetermined conversation history to the search query on the basis of the weight information set for the handling content to thereby amend the search query.
5. The information searching system according to claim 1, further comprising:
a usefulness-information acquisition unit configured to acquire usefulness information indicating whether the conversation histories included in the search result are useful; and
a search-result rearrangement unit configured to rearrange order of the conversation histories included in the search result on the basis of the usefulness information.
6. The information searching system according to claim 5, wherein a value of the usefulness information is calculated according to attributes of a user who sets the usefulness information for the conversation histories.
7. The information searching system according to claim 2, wherein the searching unit is configured to set, as a search target, a conversation history text file obtained by collecting, as one file, conversation histories having a same matter number among the conversation histories stored in the conversation-history storage unit.
8. The information searching system according to claim 2, wherein the searching unit is configured to output, as the search result, a part of the conversation histories selected out of the conversation histories having the same matter number among the conversation histories stored in the conversation-history storage unit.
9. A method of searching information using a computer, the method comprising:
acquiring, from a conversation-history storage unit configured to store conversation histories for inquiries, a predetermined conversation history corresponding to a search query based on input information;
amending the search query on the basis of the acquired predetermined conversation history;
searching, on the basis of the amended search query, the conversation histories stored in the conversation-history storage unit; and
outputting a search result.
US16/140,683 2018-02-09 2018-09-25 Information searching system and information searching method Abandoned US20190251201A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2018-021605 2018-02-09
JP2018021605 2018-02-09

Publications (1)

Publication Number Publication Date
US20190251201A1 true US20190251201A1 (en) 2019-08-15

Family

ID=67541721

Family Applications (1)

Application Number Title Priority Date Filing Date
US16/140,683 Abandoned US20190251201A1 (en) 2018-02-09 2018-09-25 Information searching system and information searching method

Country Status (2)

Country Link
US (1) US20190251201A1 (en)
JP (1) JP6832903B2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190370035A1 (en) * 2018-05-29 2019-12-05 Wipro Limited Method and system for providing real-time guidance to users during troubleshooting of devices
US11461417B2 (en) 2020-03-05 2022-10-04 Kore.Ai, Inc. Methods for managing automated discovery and knowledge sharing in one or more networks and devices thereof

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20220025860A (en) 2019-07-30 2022-03-03 토소가부시키가이샤 Cyclic azine compound, material for organic electroluminescent device, electron transport material for organic electroluminescent device, and organic electroluminescent device

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2002278977A (en) * 2001-03-22 2002-09-27 Fujitsu Ltd Device and method for answering question and question answer program
JP2002334146A (en) * 2001-05-08 2002-11-22 Mitsubishi Electric Corp Device and program for supporting creation of procedure manual
JP2010282403A (en) * 2009-06-04 2010-12-16 Kansai Electric Power Co Inc:The Document retrieval method
WO2015181962A1 (en) * 2014-05-30 2015-12-03 株式会社日立製作所 Search method and device, and storage medium
JP6439434B2 (en) * 2014-12-18 2018-12-19 富士通株式会社 Knowledge extraction editing program, knowledge extraction editing method, knowledge extraction editing apparatus, and knowledge extraction editing system

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190370035A1 (en) * 2018-05-29 2019-12-05 Wipro Limited Method and system for providing real-time guidance to users during troubleshooting of devices
US11010180B2 (en) * 2018-05-29 2021-05-18 Wipro Limited Method and system for providing real-time guidance to users during troubleshooting of devices
US11461417B2 (en) 2020-03-05 2022-10-04 Kore.Ai, Inc. Methods for managing automated discovery and knowledge sharing in one or more networks and devices thereof

Also Published As

Publication number Publication date
JP2019139746A (en) 2019-08-22
JP6832903B2 (en) 2021-02-24

Similar Documents

Publication Publication Date Title
US8370334B2 (en) Dynamic updating of display and ranking for search results
US20190349320A1 (en) System and method for automatically responding to user requests
US9973457B2 (en) Method and apparatus for live chat integration
US9218568B2 (en) Disambiguating data using contextual and historical information
US20170103439A1 (en) Searching Evidence to Recommend Organizations
KR20160021110A (en) Text matching device and method, and text classification device and method
CN103493045A (en) Automated answers to online questions
US20190251201A1 (en) Information searching system and information searching method
US11269894B2 (en) Topic-specific reputation scoring and topic-specific endorsement notifications in a collaboration tool
US11755663B2 (en) Search activity prediction
US20200034481A1 (en) Language agnostic data insight handling for user application data
JP2019121392A (en) Information search system
US9619534B2 (en) Probabilistic tree-structured learning system for extracting contact data from quotes
US20200342165A1 (en) Management of annotation jobs
CN114003818A (en) Page recommendation method, device, server and storage medium
US20110289070A1 (en) Dynamic resource orchestration system for data retrieval and output generation
US10810273B2 (en) Auto identification and mapping of functional attributes from visual representation
JP2019204157A (en) Inquiry apparatus specification system, inquiry apparatus specification method
US20230334072A1 (en) Automatic question answering system and question-answer pair data generation method
CN111126073B (en) Semantic retrieval method and device
Qumsiyeh et al. Searching web documents using a summarization approach
JPH10207902A (en) Information service system and information retrieving method
JP2011170535A (en) Document quality evaluation system and document quality evaluation program
CN113032436B (en) Searching method and device based on article content and title
JP2020067864A (en) Knowledge search device, method for searching for knowledge, and knowledge search program

Legal Events

Date Code Title Description
AS Assignment

Owner name: HITACHI, LTD., JAPAN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:CHIGUSA, KENTARO;TSUCHIDA, MASASHI;REEL/FRAME:046958/0625

Effective date: 20180731

STPP Information on status: patent application and granting procedure in general

Free format text: NON FINAL ACTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: FINAL REJECTION MAILED

STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION